0.47.0rc4

What's Changed

• fix: ping numpy <2.0 by @ThibaultFy in #217
• chore: allow skip pure torch computation on benchmark camelyon by @ThibaultFy in #219
• chore(deps): bump tqdm from 4.62.3 to 4.66.3 in /benchmark/camelyon by @dependabot in #212
• chore: add cancel cp option to camelyon benchmark by @ThibaultFy in #218
• chore(deps): bump scikit-learn from 1.3.1 to 1.5.0 in /benchmark/camelyon by @dependabot in #216
• chore: cancel cp to bool by @ThibaultFy in #220
• chore(benchmark): don't wait if the cp is canceled by @ThibaultFy in #221
• fix: mark simulation mode tests as subprocess-only by @SdgJlbl in #229
• fix: ensure that compute functions are running as non-root by @SdgJlbl in #228
• chore: support Python 3.12 by @SdgJlbl in #226
• chore: bump Pytorch version to 2.2.1 in tests by @SdgJlbl in #230
• chore: bump NumPy by @SdgJlbl in #231
• feat: towncrier reusable workflow by @ThibaultFy in #234
• chore: rename use_gpu to disable_gpu in TorchAlgo by @ThibaultFy in #241
• chore: gpu base docker image by @ThibaultFy in #227
• fix(benchmark/camelyon): actually rely on gpu setup at Dependency level by @thbcmlowk in #244
• fix(camelyon/weldon_fedavg): can't convert cuda:0 device type tensor to numpy by @thbcmlowk in #245

New Contributors

• @dependabot made their first contribution in #212

Full Changelog: 0.46.0...0.47.0rc4